There are many biological constraints on learning and the associability of various events. For example, when a tone-plus-light compound is food related, light is selectively attended to, but the tone gains considerable control when the compound is associated with shock. It has not been recognized that in these studies reporting stimulus-reinforcer interactions, food and shock were confounded with schedule component preference, since food is a preferred condition, and shock nonpreferred. A pilot study measured selective associations for the first time in a paradigm designed to unconfound preference and class-of-reinforcer. Component preferences were generated solely with differential positive reinforcement. The interaction profile generated in that experiment was like those produced when selective associations to shock and to food related compounds were compared. The proposed research program is based on the assumption that many of what up to now have been classified as stimulus-reinforcer interactions might be special instances of a stimulus-preference interaction, with the latter interactions subsuming the former. The proposed research would systematically test this hypothesis parametrically, with different species, and in situations where differential preferences are generated solely with negative reinforcement contingencies. The stimulus-preference interaction model parsimoniously relates these selective associations to fundamental psychological processes responsible for choice behavior and conditioned preference.